Online slots and casino game data, minus the sales pitch

How to read an RTP figure before you trust it

Return to player is the number most people look for first, and it is also the one most often misread. RTP is a long-run average produced by the game’s mathematical model over millions of rounds: a slot published at 96.50% is designed to return 96.50 of every 100 units staked across that whole volume, and to keep the other 3.50 as the house edge. It says nothing about a single evening, and it is only true for the specific version of the game that is actually loaded in front of you. Two players on the same title at two different casinos can be playing two different mathematical products without any visible difference on screen.

One title, several certified versions

Most large studios certify the same game at several return levels and let each operator choose one. Play’n GO, for example, has described certifying games at five tiers, roughly 96, 94, 91, 87 and 84 per cent, which is why Book of Dead can return 96.21% on one site and 84.18% on another while looking identical. Pragmatic Play publishes reduced versions of titles such as Big Bass Bonanza as well. The figure printed on a review page, including on this one, is the developer’s headline configuration, not a promise about the copy a given casino runs.

Where to find the active figure in the game panel

The only figure that applies to your session is the one inside the game itself. Regulated markets such as Great Britain require it to be available to the player, and most studios display it in the same place, close to the end of the rules screen. The check takes less than a minute and follows the same order in almost every modern title:

1. open the game and wait for the lobby or base screen to finish loading;
2. tap the information, help or paytable icon, usually marked with an “i” or a menu symbol;
3. scroll to the end of the rules, where the theoretical return is normally stated;
4. check whether a separate figure is listed for bonus buy, ante bet or side bets;
5. compare that number with the developer’s headline figure on the sheet in this index.

If the in-game number is lower than the headline, the operator has chosen a reduced configuration. That is legal in most jurisdictions, but it changes the cost of playing, and a gap of five or ten points is far larger than any difference in design between two popular online slots. Over a thousand one-unit spins, the difference between 96% and 91% is fifty units of expected cost.

Volatility, hit frequency and maximum win in online slots

Two games with the same RTP can behave in completely different ways. Volatility describes how the return is distributed: a low-volatility game pays small amounts often, while a high-volatility game holds most of its return back for rare, large outcomes. Every developer uses its own scale, which is why this index reports the studio’s own classification rather than inventing a common score, and why the maximum win is shown next to it on every sheet.

Why the cap matters more than the headline

The maximum win is a ceiling, not an expectation. Money Train 3 caps a single round at 100,000x the stake and Fire in the Hole xBomb at 60,000x, while Big Bass Bonanza stops at 2,100x. A higher cap usually means that more of the return is stored in outcomes that most players will never see, so the ordinary spins have to pay less. That is the real meaning of “very high volatility”: the published return is the same, but far more of it sits in the extreme tail of the distribution, and long runs without a meaningful win are part of the design rather than a sign that a game is “due”.

What 200 spins actually cost

A simple calculation makes the figures concrete. Two hundred spins at one unit each on a game published at 96.50% put 200 units into play, and the expected cost of that volume is 3.5% of it, or seven units. The result of any real session will almost never be seven units down: in a high-volatility slot it is far more likely to be a loss of most of the budget or, much less often, a large win. The expected cost only becomes visible over a very long run, which is exactly why it is easy to forget during a short one.

Bonus buy, ante bets and other features that change the cost

Many modern online slots offer ways to change the stake or skip straight to the bonus round, and each one has its own published return. The headline figure on a sheet normally describes standard play, so the first question about any optional feature is whether the studio publishes a separate number for it, and whether that number is higher, lower or identical to the base game.

When buying the bonus changes the RTP

A feature buy lets the player pay a fixed multiple of the stake to enter the bonus round at once. On most titles the return of the buy is close to the base figure, but there are exceptions in both directions. Money Train 3 is a rare case where every one of its four feature buys returns 96.50% against 96.10% in standard play. Feature buys are also restricted in several markets: in Great Britain they are not permitted, so the option simply does not appear at UKGC-licensed operators, whatever a review page says.

Ante bets and higher stakes

An ante bet raises the stake, typically by around a quarter, in exchange for more frequent bonus triggers. On Gates of Olympus 1000, Pragmatic Play lists the same 96.50% for standard and ante play, so the option changes the cost per spin and the rhythm of the session without improving the mathematical position. Reaching the free spins more often on a larger stake is a different session, not a better one, and the expected cost per hour rises in proportion to the extra stake, even though the published percentage stays exactly where it was.

Crash games and instant games on the same maths

Crash games such as Aviator, JetX and Spaceman, and instant games such as Mines and Plinko, look nothing like a slot, but they rest on the same idea: a random outcome generated before or at the start of the round, and a published return that holds over a very large number of rounds. Most of the crash titles in this index publish 97.00%, and Chicken Road publishes 98.00%. Rounds are short, often under a minute, so the volume staked per hour can be much higher than in a slot played at the same stake, and the expected cost grows with that volume.

Cash-out decisions and the published return

The cash-out button gives a strong sense of control, and that sense is the most common misunderstanding about these crash games. Cashing out early or late changes how results are spread, more small wins or fewer large ones, but it does not change the long-run return, which is fixed by the crash-point distribution. Auto cash-out at a target multiplier is a convenience, not a strategy, and no sequence of past rounds tells you anything about the next one. Every round starts from the same odds, however long the previous streak of low crash points has been.

What provably fair claims verify

Several studios, Spribe among them, describe their crash games as provably fair: the outcome of a round is derived from seeds that can be checked after it ends. That mechanism proves the result was not changed after bets were placed. It does not prove that the return is generous, that the operator runs the headline configuration, or that the site is licensed, so it should be read as one technical safeguard rather than as a reason to trust a casino. Licensing, the operator’s complaint procedure and the in-game RTP still need to be checked separately.

Live casino and game shows

Live casino covers two very different products. Classic tables streamed from a studio, such as Live Blackjack and Lightning Roulette, follow rules close to their land-based versions. Game shows such as Crazy Time and MONOPOLY Big Baller mix a physical wheel or ball draw with bonus rounds and multipliers, and their returns are lower and more spread out. The stake limits, the number of seats and the bonus rules can also vary between studios and markets, so the rules screen of each table remains the reference.

Tables where decisions change the return

Blackjack is the clearest case of a game in which the player’s choices matter. The 99.29% published for Live Blackjack assumes optimal basic strategy on every hand; each departure from it, such as taking insurance or standing on a weak total against a strong dealer card, lowers the return actually received. Roulette offers no such lever: every bet on a single-zero wheel returns 97.30%, except where a variant such as Lightning Roulette deliberately trades a lower base payout for rare multipliers.

Game shows where the bet spot sets the return

In game shows the return depends on where the stake is placed rather than on any decision during the round. Crazy Time is the best-known example: the most common number on the wheel carries the highest figure, around 96.08%, while the bonus game spots are priced lower, down to about 94.33% for Pachinko. Spreading stakes across every spot to “catch” a bonus therefore lowers the average return of the session rather than raising it, and the Top Slot multipliers that make the show exciting are already built into those figures.

Highest and lowest published returns in this index

Across the 32 games documented here, published returns range from 99.29% to 94.33%, and the pattern is consistent: table games with simple rules sit at the top, crash and instant games cluster around 97%, most modern online slots land between 96% and 97%, and game-show bonus spots sit at the bottom. The table summarises the extremes of each category, using the developer’s headline figure for every title.

A higher figure is not automatically a better choice. Live Blackjack only reaches 99.29% when every hand is played by basic strategy, Chicken Road’s 98% applies to the developer’s headline configuration, and a slot at 96.86% can be far more volatile than one at 96%. The figures are most useful for comparing versions of the same game, or for spotting a reduced configuration in the game panel, rather than for ranking titles against each other.

Who makes the game and who runs the casino

A game passes through several companies before it reaches a screen, and each one is responsible for something different. Knowing which is which explains why a complaint about a withdrawal goes to one company and a question about a game’s mathematics to another. The main roles in the supply chain of online casino games are these:

• the studio, which designs the game and certifies its mathematical model and return levels;
• the aggregator or platform, which distributes many studios’ games to many casinos;
• the operator, which holds the player account, chooses the configuration and handles deposits and withdrawals;
• the regulator, which licenses the operator for a specific market and enforces its rules.

A supplier licence held by a studio says nothing about the operator that shows its games. The licensing and regulation page explains how to check an operator in the public registers of the UK Gambling Commission and the Malta Gaming Authority, and every casino game providers profile in this index lists the licences each studio declares. When a figure on a sheet and the game panel disagree, the game panel wins; when an operator and a studio disagree about a payout, the operator’s licence and its regulator decide the complaint.

Using this index responsibly

Every sheet here describes how a game is built, not how to win at it, because no figure in this index can be turned into a reliable profit. Casino games are paid entertainment with a cost set by the house edge, and the cost grows with the volume staked. Setting a budget before playing, using deposit and time limits, and never chasing losses are the habits that matter; the responsible gambling page lists the tools and helplines available, including self-exclusion schemes such as GAMSTOP in Great Britain.

The warning signs are the same whatever the game: playing for longer or with more money than planned, returning quickly to win back a loss, borrowing or using money meant for bills, and hiding how much time or money goes into gambling. Anyone who recognises several of them should pause, talk to someone they trust and contact a support service; GamCare runs a free helpline in Great Britain, and most licensed operators must also offer cooling-off periods and self-exclusion from the account settings.